Clinical Characteristics of Lower‐Extremity Edema in Stage A Cardiovascular Disease Status Defined by the ACC/AHA 2001 Chronic Heart Failure Guidelines

Abstract

Background

Checking for lower‐extremity edema is important for diagnosing, monitoring, and managing heart failure (HF). However, the characteristics of this sign in the early stages of cardiovascular disease (stage A, as defined by the American College of Cardiology/American Heart Association 2001 chronic HF guidelines) have not been adequately explored.

Hypothesis

We hypothesized that stage A HF patients (at risk for HF) are free from leg edema.

Methods

After the exclusion of patients with high serum creatinine levels (≥1.2 mg/dL), abnormal electrocardiogram rhythm, and/or comorbid critical disease(s), we studied the prevalence and predictors for leg edema in 274 subjects at risk for heart failure but without structural heart disease or symptoms of HF.

Results

The overall prevalence of lower‐extremity edema was 33 of 274 patients (12.0%; 95% confidence interval: 8.2%‐15.9%). In most patients (29/33, 88%), the leg edema involved only the ankle and foot. Compared with patients without leg edema (n = 241), those with leg edema (n = 33) were older (age, 74 ± 11 7 vs 69 ± 12 years, P = 0.006), more likely to present with pulmonary crackles (52% vs 31%, P < 0.03), and more likely to have varicose veins (55% vs 15%, P < 0.001). Leg edema and varicose veins coexisted in 19 (58%) patients, of whom 16 patients were women. On multivariate analysis, the presence of varicose veins was the only independent predictor of the appearance of bilateral leg edema (odds ratio: 8.18, 95% confidence interval: 3.92‐17.1, P < 0.001).

Conclusions

A mild degree of leg edema is not uncommon in patients at risk for HF. Recognition of this phenomenon might be important for evaluation, monitoring, and self‐care of HF patients.

Introduction

Signs of heart failure (HF) include right‐heart congestion (eg, lower‐leg edema, raised jugular venous pressure, ascites, and/or hepatomegaly) and left‐heart congestion (pulmonary crackles or wheeze).1, 2, 3 Among these signs of congestion in HF patients, lower‐extremity edema is one of the important factors in diagnosing, monitoring, and managing HF status.4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 Lower‐extremity edema is a common physical finding with numerous etiologies,19, 20, 21 including venous insufficiency,22 adverse reaction to medication,23, 24, 25 chronic obstructive pulmonary disease,26, 27 sleep apnea or hypopnea syndrome,28, 29, 30 obesity,29, 30, 31 and elevated central venous pressure due to HF or renal failure.32, 33 The presence of frequent noncardiac causes of leg edema might interfere with the physician's management of patients with suspected HF, particularly in the early stages of cardiovascular disease (CVD). Little is known about these issues in patients with early‐stage CVD. Because diagnostic and therapeutic decisions may be made on the basis of lower‐extremity edema, knowledge of its prevalence among patients at risk for HF with CVD could be clinically important. Therefore, the present study examined the characteristics of lower‐extremity edema among adult patients with stage A CVD status as defined by the American College of Cardiology/American Heart Association 2001 chronic HF guidelines (ie, patients at high risk for congestive HF, but without structural heart disease or symptoms of HF).34

Methods

Study Design and Patients

The present study is a substudy of a previously published report35 focusing on the appearance of pulmonary crackles in stage A CVD status as defined by the presence of HF risk factors (ie, hypertension, diabetes mellitus, obesity [defined as body mass index ≥30 kg/m²], and coronary artery disease) without cardiac structural or functional abnormality,34 which was performed in the cardiology outpatient clinic of Nishida Hospital (Oita, Japan). The protocol was approved by ethics committee of Nishida Hospital, and all patients provided informed consent.

Adult outpatients (age ≥45 years) undergoing treatment for cardiovascular diseases, including metabolic syndrome, were first screened by research nurses, who obtained consent from the patients and recorded a brief cardiac and pulmonary history and current complaints. Patients were not eligible for the present study if acute cardiopulmonary complaints were indicated or if they had a current or previous diagnosis of structural heart diseases, decompensated HF, chronic pulmonary disease, and/or a recent episode of acute respiratory disease.

All screened patients underwent a physical examination including the measurement of body height and weight, blood chemistry tests, a 12‐lead electrocardiogram (ECG), and a simple chest x‐ray. At this second step, patients were excluded if they had high serum creatinine levels (≥1.2 mg/dL), ECG abnormalities (atrial fibrillation and/or bundle branch block), or remarkable chest x‐ray abnormalities. Finally, all the patients screened at the second step underwent a cardiac echo/Doppler study and blood sampling for analysis of serum B‐type natriuretic peptide (BNP) levels.36 At this final step, patients were excluded if they had abnormal echo/Doppler results37, 38 or high BNP levels (≥80 pg/mL).39, 40

A total of 255 patients were followed up for periods ranging from 6 to 12 months. During this follow‐up period, the occurrence of critical cardiovascular diseases was evaluated.

Evaluation of Leg Edema

A single senior cardiologist (H.K.) checked carefully for the presence of bilateral lower‐extremity edema and graded its severity on a scale of 1 of 4 based on the extent of the edema in the leg, as follows: (1) none, (2) up to the height of the ankle level (mild edema), (3) up to the height of the knee (moderate edema), and (4) higher than knee level (severe edema).

Also, the presence or absence of varicose veins was determined by physical examination.22 With the patient in the upright posture to allow for maximal distention of the veins from multiple directions, inspection and palpation were performed to detect evidence of chronic venous disease, as such examining the skin surface for irregularities or bulges suggesting the presence of varicose veins.

Other Measurements

Body mass index was derived from the body height and weight, and obesity was defined as body mass index ≥30 kg/m².41 Serum BNP was measured using the Shionoria assay method.36 An echo/Doppler study was performed using a commercially available real‐time, wide‐angle, phased‐array system (Aloka SSD‐2000; Aloka Co. Ltd., Tokyo, Japan).

Statistical Analysis

Continuous variables are expressed as mean values ± standard deviation and categorical data as frequencies and percentages. Patients were categorized into 2 groups based on the presence or absence of bilateral leg edema. Continuous variables were examined using a 2‐tailed unpaired Student t test. Categorical variables were compared using the χ² test or the Fisher exact test. A P value of 0.05 was considered statistically significant. Backward logistic regression analysis was used to determine the predictors of bilateral leg edema by taking univariate predictors of the appearance of leg edema and using iterative modeling procedures to arrive at the most efficient model. The threshold for entry of variables into the model was P < 0.20. The odds ratio (OR) and associated 95% confidence interval (CI) were estimated to determine the association between those variables and the presence of leg edema.

Results

The Figure 1 shows the flow diagram describing the selection of the study patients for the present analysis. Of a total of 385 study patients under current treatment for hypertension, diabetes mellitus, and/or dyslipidemia, 52 patients were excluded because of high serum creatinine levels (n = 7), ECG abnormalities (n = 13), major chest x‐ray abnormality (n = 15), or any combination of these (n = 17). The remaining 333 patients underwent a cardiac echo/Doppler study and measurement of serum BNP levels. At this step, 59 patients were excluded from the analysis due to abnormal echo/Doppler results (n = 11), high BNP levels (n = 14), or both (n = 34). After this final exclusion, 274 asymptomatic adult patients remained eligible because they were in sinus rhythm and they had normal echocardiograms, BNPs, and chest radiographs.

Figure 1.

Flow diagram describing the steps of inclusion and exclusion of study patients for the analysis. Abbreviations: ECG, electrocardiogram.

The overall prevalence of lower‐extremity edema was 33 of 274 patients (12.0%; 95% CI: 8.2%‐15.9%). In most patients (29/33, 88%), the extent of the leg edema was lower than or equal to ankle height. In the present study, no patients had unilateral leg edema, but 3 had asymmetric bilateral leg edema. Characteristics of patients with bilateral leg edema and those without are presented in Table 1. Compared with patients without leg edema (n = 241), those with leg edema (n = 33) were older (age, 74 ± 11 vs 69 ± 12 years, P = 0.006), presented with pulmonary crackles (52% vs 31%, P < 0.03), and had visible varicose veins (55% vs 15%, P < 0.001). Leg edema and varicose veins coexisted in 19 (58%) patients, of whom 16 patients were women.

Table 1.

Characteristics of the Study Groups

Variable	All Patients, N = 274	Bilateral Leg Edema		P Value
		Presence, n = 33	Absence, n = 241
Age, y	69.2 ± 12	74.4 ± 10.7	68.5 ± 11.5	0.006
Sex, M/F	77/197	6/27	71/170	0.18
Obesity (BMI ≥30 kg/m2)	27 (10)	5 (15)	22 (9)	0.34
Treatment
Antihypertensive agents
ACEI/ARB	98 (36)	13 (39)	85 (35)	0.67
β‐Blocker	104 (38)	13 (39)	91 (38)	0.85
CCB	168 (61)	20 (61)	148 (61)	1
Diuretics	37 (14)	8 (24)	29 (12)	0.1
Lipid‐lowering agents	88 (32)	10 (30)	78 (32)	1
Antiarrhythmics	39 (14)	5 (15)	34 (14)	0.79
Nitrates	28 (10)	5 (15)	23 (10)	0.35
DM	24 (9)	5 (15)	19 (8)	0.18
Audible pulmonary crackles	92 (34)	17 (52)	75 (31)	0.03
Third heart sound	4 (1.5)	0	4 (1.7)	1
Varicose veins in the legs	58 (21)	19 (55)	39 (15)	<0.001
LVEF, %	61.3 ± 7.1	61.5 ± 7.0	61.3 ± 7.1	0.92
Serum Cr, mg/dL	0.77 ± 0.21	0.78 ± 0.21	0.76 ± 0.21	0.68
BNP, pg/mL	31.6 ± 19.4	33.7 ± 18.7	31.3 ± 19.6	0.5

Abbreviations: ACEI, angiotensin‐converting enzyme inhibitor; ARB, angiotensin II receptor blocker; BMI, body mass index; BNP, B‐type natriuretic peptide; CCB, calcium channel blocker; Cr, creatinine; DM, diabetes mellitus; F, female; LVEF, left ventricular ejection fraction; M, male.

Data are given as n (%) or mean ± standard deviation.

Results of the multivariate logistic regression analysis are presented in Table 2. The presence of varicose veins was the only independent predictor of bilateral leg edema (OR: 8.18, 95% CI: 3.92‐17.1, P < 0.001). The other variables examined were not significant predictors of leg edema, although there was a trend toward an independent association between the appearance of leg edema and higher age (OR: 2.24, 95% CI: 0.97‐5.17, P < 0.06).

Table 2.

Multivariate Predictors of Bilateral Leg Edema

Variable	OR of Leg	95% CI	P Valuea
Age (≥72 y)	2.24	0.97–5.17	0.06
Sex (F)	0.826	0.346–1.98	0.67
Use of diuretics	1.48	0.587–3.74	0.41
Treatment for DM	1.72	0.587–3.74	0.33
Audible pulmonary crackles	1.25	0.571–2.74	0.58
Varicose veins in the leg	8.18	3.92–17.1	<0.001

Abbreviations: CI, confidence interval; DM, diabetes mellitus; F, female; OR, odds ratio.

^aBackward logistic regression analysis including the variables of P < 0.02 by univariate analysis.

During the follow‐up of a mean of 11 ± 2.3 months in 255 patients, new‐onset HF occurred in 3 patients, 1 with and 2 without leg edema (not significant).

Discussion

The currently available evidence indicates that HF is usually a progressive condition that begins with risk factors for cardiac dysfunction, proceeds to asymptomatic changes in cardiac structure and function, and then evolves into clinically overt HF, disability, and death.34 In recognition of the importance of this concept, a joint statement from the American College of Cardiology/American Heart Association proposed a novel HF model that classifies HF into 4 stages from A to D: stage A, HF risk factors; stage B, asymptomatic cardiac structural or functional abnormalities; stage C, symptomatic HF; and stage D, end‐stage HF.34 In a recent cohort study by Ammar et al,42 the prevalence of the proposed HF stages in the community was reported to be 22% in stage A, 34% in stage B, 12% in stage C, and 0.2% in stage D. To date, studies of the clinical pictures of stage A HF status are still lacking, including the report by Ammar et al.42 The present study is the first to examine the appearance of bilateral lower‐extremity edema in stage A HF status, and the results demonstrated that the incidence of bilateral lower‐extremity edema was approximately 10% in this HF population subset. In most of the patients with leg edema, the severity was mild (eg, limited to the ankle and foot). An independent factor associated with the appearance of lower‐extremity edema was the presence of varicose veins in the leg; another, with borderline statistical significance, was age.

According to recent studies evaluating HF‐related physical findings, Mueller et al14 observed the appearance of lower‐extremity edema in a moderate percentage of patients with a noncardiac cause of dyspnea (55/235 patients, 23%) compared with those with a cardiac cause of dyspnea (101/217 patients, 47%) in an emergency‐department setting. Kelder et al16 reported a similar observation in the primary‐care setting for diagnosing new‐onset HF: the incidence of bilateral ankle edema was 113 of 514 patients (22%) without HF vs 84 of 207 patients (40.6%) with HF. In an assessment of geriatric patients with suspected HF, Oudejans et al17 reported that the appearance of bilateral ankle swelling was not different between those diagnosed as not having HF (40 of 112 patients, 36%) and those having new, slow‐onset HF (40 of 94 patients, 43%). Damy et al15 noted that a third of patients with ankle swelling alone, and a similar proportion of those with lung crackles alone, did not have HF. Overall, these recent studies, including the present study, assessing real‐world populations with suspected HF, reconfirmed the previous observations4, 5, 6, 7, 8, 9 indicating that the use of lower‐leg edema as a predictor of HF is the simplest sign of right‐sided congestion, but it is not specific enough to use as the sole basis for an accurate diagnosis of worsening HF status. Diagnosis and monitoring of HF is difficult, partly because of the nebulous concept of continued assessment of a complex and ill‐defined disease process that is associated with significant disability and mortality.43 The researchers, however, used all possible information to make the final adjudicated diagnosis of worsening HF. A combination of ≥2 physical signs of right‐sided (eg, jugular venous distension) and left‐sided congestion (eg, lung crackles), if present, could provide higher predictive value of the presence of HF, despite the lack of sensitivity for its identification.11, 15 Also, the addition of clinical tests to the physical signs could improve the ability to diagnose worsened HF, such as measurement of serum BNP.16, 17

Study Limitations

The present study had certain limitations. First, the criteria to assess leg edema might somewhat affect the grading of the edema. The severity of edema, from slight to very marked, is traditionally reported on a 4‐point scale based on the depth of indentation at the ankle.14 Because this scale is subjective, noting the height of the edema may provide more practical and reproducible information and is currently used universally, as in the present study.20 Second, the mean BNP value in patients assigned to stage A HF status in the present study was similar to that of the cohort study by Ammar et al42 (32 pg/mL [95% CI: 25‐40 pg/mL] vs 32 pg/mL [95% CI: 29‐34 pg/mL]), but the distribution of patient age between these 2 studies was significantly different, with more older patients (age ≥65 years) in the present study (177 of 274 patients, 65.0%) compared with the cohort study by Ammar et al42 (130 of 454 patients, 28.6%; P < 0.0001). Such a selection difference might limit the generalization of the clinical pictures observed in the present study to other populations assigned to stage A HF status, because the sensitivity and specificity of some of the HF‐related signs and symptoms might be age‐dependent. Indeed, as for lower‐leg edema, Mueller et al14 demonstrated that its appearance is more sensitive, but less specific, in elderly HF patients. The present study supports these findings, because leg edema often occurred in older participants without HF, which could have worsened the specificity of this HF‐related sign for diagnosing truly worsening HF status in the elderly population. Third, the present study confirmed previous observations indicating an association between the appearance of edema and venous insufficiency in the lower‐leg edema.19, 20, 21, 22 In the present study, we determined the presence of venous insufficiency solely by the presence of varicose veins upon physical examination. However, leg edema from venous insufficiency can occur in the absence of varicose veins, and the varicose veins can occur in the absence of leg edema. So, it might be necessary to clarify the associations among varicose veins, venous insufficiency, and leg edema. The utility of duplex imaging and/or continuous‐wave Doppler might provide more adequate information to help understand the contribution of venous insufficiency to the appearance of leg edema.22, 44 Fourth, although obesity is frequently associated with lower‐extremity edema,29, 30, 31 findings from the present study did not confirm this association in patients with stage A CVD status. Because of the small sample, however, a type II error might explain the lack of a difference in the appearance of obesity between those with positive or negative lower‐extremity edema. Alternatively, severely obese patients, in whom a high possibility of accompanying leg edema could be expected,30 might have been excluded from the present study during the selection process of patients with stage A CVD status because severely obese patients might have had a high probability of cardiac structural or functional abnormalities.31 Finally, direct measurements of right and left heart filling pressures might provide valuable information to elucidate the pathophysiology of leg edema, but these data are not available due to the lack of rationale for performing such an invasive examination in the present study population.

Conclusion

A mild degree of leg edema is not uncommon in patients at risk for HF. Recognition of this phenomenon might be important for evaluation, monitoring, and self‐care of HF patients.